[PATCH 2/3] libnvdimm: Add a device-tree interface

Rob Herring robh at kernel.org
Fri Nov 17 01:28:40 AEDT 2017


On Thu, Nov 16, 2017 at 04:51:30AM +1100, Oliver O'Halloran wrote:
> A fairly bare-bones set of device-tree bindings so libnvdimm can be used
> on powerpc and other device-tree based platforms.
> 
> Cc: devicetree at vger.kernel.org
> Signed-off-by: Oliver O'Halloran <oohall at gmail.com>
> ---
>  .../devicetree/bindings/nvdimm/nvdimm-bus.txt      |  69 +++++++
>  MAINTAINERS                                        |   8 +
>  drivers/nvdimm/Kconfig                             |  10 +
>  drivers/nvdimm/Makefile                            |   1 +
>  drivers/nvdimm/of_nvdimm.c                         | 202 +++++++++++++++++++++
>  5 files changed, 290 insertions(+)
>  create mode 100644 Documentation/devicetree/bindings/nvdimm/nvdimm-bus.txt
>  create mode 100644 drivers/nvdimm/of_nvdimm.c
> 
> diff --git a/Documentation/devicetree/bindings/nvdimm/nvdimm-bus.txt b/Documentation/devicetree/bindings/nvdimm/nvdimm-bus.txt
> new file mode 100644
> index 000000000000..491e7c4900ed
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/nvdimm/nvdimm-bus.txt
> @@ -0,0 +1,69 @@
> +Device-tree bindings for nonvolatile-memory / NVDIMMs
> +-----------------------------------------------------
> +
> +Non-volatile DIMMs are memory modules used to provide (cacheable) main memory that
> +retains its contents across power cycles. In more practical terms, they are kind
> +of storage device where the contents can be accessed by the CPU directly,
> +rather than indirectly via a storage controller or similar. This can provide
> +substantial performance improvements for applications designed to take
> +advantage of in-memory storage.
> +
> +This binding provides a way to describe memory regions that should be managed
> +by an NVDIMM storage driver (libNVDIMM in Linux) and some of the associated
> +metadata. The binding itself is split into two main parts: A container bus and
> +its sub-nodes which describe which memory address ranges corresponding to
> +NVDIMM backed memory.
> +
> +Bindings for the container bus:
> +------------------------------
> +
> +Required properties:
> + - compatible = "nvdimm-bus";
> + - ranges;
> +	A blank ranges property is required because the sub-nodes have
> +	addresses in the system's physical address space.
> +
> +The use of a container bus is mainly to handle future expansion of the binding. For
> +comparison the ACPI equivalent of this binding (NFIT) describes: Memory regions, DIMM
> +control structures, Block mode DIMM control structures, interleave sets, and more. Some
> +of these structures cross reference each other so everyone should be happier if we keep
> +it relatively self contained.

Will adding any of these things need unit addresses and colide with the 
existing nodes below? IOW, at one level there's only 1 number space.

> +
> +Bindings for the region nodes:
> +-----------------------------
> +
> +Required properties:
> +	- compatible = "nvdimm-persistent" or "nvdimm-volatile"
> +
> +		The "nvdimm-persistent" region type indicates that this memory region
> +		is actually a persistent region. The volatile type is mainly useful
> +		for testing and RAM disks that can persist across kexec.
> +
> +	- reg = <base, size>;
> +		The reg property should only contain one address range.
> +
> +Optional properties:
> +	- Any relevant NUMA assocativity properties for the target platform.
> +
> +A complete example:
> +--------------------
> +
> +/ {
> +	#size-cells = <1>;
> +	#address-cells = <1>;
> +
> +	nonvolatile-memory {
> +		compatible = "nonvolatile-memory";
> +		ranges;
> +
> +		region at 5000 {
> +			compatible = "nvdimm-persistent";
> +			reg = <0x5000 0x1000>;
> +		};
> +
> +		region at 6000 {
> +			compatible = "nvdimm-volatile";
> +			reg = <0x6000 0x1000>;
> +		};
> +	};
> +};
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 65eff7857ec3..0350bf5a94d2 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -7875,6 +7875,14 @@ Q:	https://patchwork.kernel.org/project/linux-nvdimm/list/
>  S:	Supported
>  F:	drivers/nvdimm/pmem*
>  
> +LIBNVDIMM: DEVICETREE BINDINGS
> +M:	Oliver O'Halloran <oohall at gmail.com>
> +L:	linux-nvdimm at lists.01.org
> +Q:	https://patchwork.kernel.org/project/linux-nvdimm/list/
> +S:	Supported
> +F:	drivers/nvdimm/of_nvdimm.c
> +F:	Documentation/devicetree/bindings/nvdimm/nvdimm-bus.txt
> +
>  LIBNVDIMM: NON-VOLATILE MEMORY DEVICE SUBSYSTEM
>  M:	Dan Williams <dan.j.williams at intel.com>
>  L:	linux-nvdimm at lists.01.org
> diff --git a/drivers/nvdimm/Kconfig b/drivers/nvdimm/Kconfig
> index 5bdd499b5f4f..72d147b55596 100644
> --- a/drivers/nvdimm/Kconfig
> +++ b/drivers/nvdimm/Kconfig
> @@ -102,4 +102,14 @@ config NVDIMM_DAX
>  
>  	  Select Y if unsure
>  
> +config OF_NVDIMM
> +	tristate "Device-tree support for NVDIMMs"
> +	depends on OF
> +	default LIBNVDIMM
> +	help
> +	  Allows byte addressable persistent memory regions to be described in the
> +	  device-tree.
> +
> +	  Select Y if unsure.
> +
>  endif
> diff --git a/drivers/nvdimm/Makefile b/drivers/nvdimm/Makefile
> index ca6d325438a3..9029511a8486 100644
> --- a/drivers/nvdimm/Makefile
> +++ b/drivers/nvdimm/Makefile
> @@ -3,6 +3,7 @@ obj-$(CONFIG_BLK_DEV_PMEM) += nd_pmem.o
>  obj-$(CONFIG_ND_BTT) += nd_btt.o
>  obj-$(CONFIG_ND_BLK) += nd_blk.o
>  obj-$(CONFIG_X86_PMEM_LEGACY) += nd_e820.o
> +obj-$(CONFIG_OF_NVDIMM) += of_nvdimm.o
>  
>  nd_pmem-y := pmem.o
>  
> diff --git a/drivers/nvdimm/of_nvdimm.c b/drivers/nvdimm/of_nvdimm.c
> new file mode 100644
> index 000000000000..765cbbae8bcb
> --- /dev/null
> +++ b/drivers/nvdimm/of_nvdimm.c
> @@ -0,0 +1,202 @@
> +/*
> + * Copyright 2017, IBM Corporation
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or
> + * (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, you can access it online at
> + * http://www.gnu.org/licenses/gpl-2.0.html.

This can be SPDX tag.

> + */
> +
> +#define pr_fmt(fmt) "of_nvdimm: " fmt
> +
> +#include <linux/of_platform.h>
> +#include <linux/of_address.h>
> +#include <linux/libnvdimm.h>
> +#include <linux/module.h>
> +#include <linux/ioport.h>
> +#include <linux/slab.h>
> +
> +static const struct attribute_group *region_attr_groups[] = {
> +	&nd_region_attribute_group,
> +	&nd_device_attribute_group,
> +	NULL,
> +};
> +
> +static int of_nvdimm_add_byte(struct nvdimm_bus *bus, struct device_node *np)

_add_byte_region() would be more clear.

> +{
> +	struct nd_region_desc ndr_desc;
> +	struct resource temp_res;
> +	struct nd_region *region;
> +
> +	/*
> +	 * byte regions should only have one address range
> +	 */
> +	if (of_address_to_resource(np, 0, &temp_res)) {
> +		pr_warn("Unable to parse reg[0] for %pOF\n", np);
> +		return -ENXIO;
> +	}
> +
> +	pr_debug("Found %pR for %pOF\n", &temp_res, np);
> +
> +	memset(&ndr_desc, 0, sizeof(ndr_desc));
> +	ndr_desc.res = &temp_res;
> +	ndr_desc.of_node = np;
> +	ndr_desc.attr_groups = region_attr_groups;
> +#ifdef CONFIG_NUMA

I believe of_node_to_nid will do the right thing for !CONFIG_NUMA and 
this can be dropped.

> +	ndr_desc.numa_node = of_node_to_nid(np);
> +#endif
> +	set_bit(ND_REGION_PAGEMAP, &ndr_desc.flags);
> +
> +	if (of_device_is_compatible(np, "nvdimm-volatile"))

You've already matched the node once with of_match_node, don't do it 
again.

> +		region = nvdimm_volatile_region_create(bus, &ndr_desc);
> +	else
> +		region = nvdimm_pmem_region_create(bus, &ndr_desc);
> +
> +	if (!region)
> +		return -ENXIO;
> +
> +	return 0;
> +}
> +
> +static const struct of_device_id of_nvdimm_dev_types[] = {
> +	{ .compatible = "nvdimm-persistent", .data = of_nvdimm_add_byte },
> +	{ .compatible = "nvdimm-volatile", .data = of_nvdimm_add_byte },
> +	{ },
> +};
> +
> +static void of_nvdimm_parse_one(struct nvdimm_bus *bus,
> +		struct device_node *node)
> +{
> +	int (*parse_node)(struct nvdimm_bus *, struct device_node *);
> +	const struct of_device_id *match;
> +	int rc;
> +
> +	if (of_node_test_and_set_flag(node, OF_POPULATED)) {

Why not make the regions platform devices too? Then you don't need your 
own populate code.

> +		pr_debug("%pOF already parsed, skipping\n", node);
> +		return;
> +	}
> +
> +	match = of_match_node(of_nvdimm_dev_types, node);
> +	if (!match) {
> +		pr_info("No compatible match for '%pOF'\n", node);
> +		of_node_clear_flag(node, OF_POPULATED);
> +		return;
> +	}
> +
> +	of_node_get(node);
> +	parse_node = match->data;

This makes no sense. I assume you plan to add some other region type 
later?

> +	rc = parse_node(bus, node);
> +
> +	if (rc) {
> +		of_node_clear_flag(node, OF_POPULATED);
> +		of_node_put(node);
> +	}
> +
> +	pr_debug("Parsed %pOF, rc = %d\n", node, rc);
> +
> +	return;
> +}
> +
> +/*
> + * The nvdimm core refers to the bus descriptor structure at runtime
> + * so we need to keep it around. Note that that this is different to
> + * the other nd_*_desc types which are essentially containers for extra
> + * function parameters.
> + */
> +struct of_nd_private {
> +	struct nvdimm_bus_descriptor desc;
> +	struct nvdimm_bus *bus;
> +};
> +
> +static const struct attribute_group *bus_attr_groups[] = {
> +	&nvdimm_bus_attribute_group,
> +	NULL,
> +};
> +
> +static int of_nvdimm_probe(struct platform_device *pdev)
> +{
> +	struct device_node *node, *child;
> +	struct of_nd_private *priv;
> +
> +	node = dev_of_node(&pdev->dev);
> +	if (!node)
> +		return -ENXIO;
> +
> +	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
> +	if (!priv)
> +		return -ENOMEM;
> +
> +	priv->desc.attr_groups = bus_attr_groups;
> +	priv->desc.provider_name = "of_nvdimm";
> +	priv->desc.module = THIS_MODULE;
> +	priv->desc.of_node = node;
> +
> +	priv->bus = nvdimm_bus_register(&pdev->dev, &priv->desc);
> +	if (!priv->bus)
> +		goto err;
> +
> +	platform_set_drvdata(pdev, priv);
> +
> +	/* now walk the node bus and setup regions, etc */
> +	for_each_available_child_of_node(node, child)
> +		of_nvdimm_parse_one(priv->bus, child);
> +
> +	return 0;
> +
> +err:
> +	nvdimm_bus_unregister(priv->bus);
> +	kfree(priv);
> +	return -ENXIO;
> +}
> +
> +static int of_nvdimm_remove(struct platform_device *pdev)
> +{
> +	struct of_nd_private *priv = platform_get_drvdata(pdev);
> +	struct device_node *node;
> +
> +	if (!priv)
> +		return 0; /* possible? */
> +
> +	for_each_available_child_of_node(pdev->dev.of_node, node) {
> +		if (!of_node_check_flag(node, OF_POPULATED))
> +			continue;
> +
> +		of_node_clear_flag(node, OF_POPULATED);
> +		of_node_put(node);
> +		pr_debug("de-populating %s\n", node->full_name);
> +	}
> +
> +	nvdimm_bus_unregister(priv->bus);
> +	kfree(priv);
> +
> +	return 0;
> +}
> +
> +static const struct of_device_id of_nvdimm_bus_match[] = {
> +	{ .compatible = "nvdimm-bus" },
> +	{ },
> +};
> +
> +static const struct platform_driver of_nvdimm_driver = {
> +	.probe = of_nvdimm_probe,
> +	.remove = of_nvdimm_remove,
> +	.driver = {
> +		.name = "of_nvdimm",
> +		.owner = THIS_MODULE,
> +		.of_match_table = of_nvdimm_bus_match,
> +	},
> +};
> +
> +module_platform_driver(of_nvdimm_driver);
> +MODULE_DEVICE_TABLE(of, of_nvdimm_bus_match);
> +MODULE_LICENSE("GPL v2");
> +MODULE_AUTHOR("IBM Corporation");
> -- 
> 2.9.5
> 
> --
> To unsubscribe from this list: send the line "unsubscribe devicetree" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html


More information about the Linuxppc-dev mailing list